failover a controller on an equallogic array - General - Forum - Storage - Dell Community

failover a controller on an equallogic array

Storage

Storage
Gain insight and information on Dell EMC Storage products from engineers, peers and experts in the industry.

failover a controller on an equallogic array

  • I want to fail controller 0 to controller 1 on a production equallogic array (during a maintenance window). Is there a way to failover a controller without having to pull power on the controller?

    Thanks,

    Adam Feind
    Midway ISD
  • Starting with Firmware version v4.2.x and above, there is a "restart" option in the GUI (select member> member name> service tag, towards the bottom is a restart link). You can also telnet the actual member (using a IP address from one of it's physical interfaces)> login as grpadmin, and type restart. This will cause a restart of the member, and the control module will failover. Be reminded that this will momentarily, restart the array member (takes about 15-20 seconds), and any servers connected to the Group will need to be properly setup to "ride out" the iSCSI time out first (see the firmware release notes for your servers's suggested iSCSI timeout values).

    -Joe

    Social Media and Community Professional
    #IWork4Dell
    Get Support on Twitter - @dellcarespro

    Follow me on Twitter: @joesatdell 

  • Thanks for the reply - so the restart does not actually restart the "array" but the controller. Is this the best way to test a failver scenario?

    I am wanting to verify my iscsi pathways on the switches, servers, etc.

    Adam
  • Yes, This method is a good test of your iscsi pathways, switch ISL's, server MPIO/multipath setup as you listed.

    Just remember that when you restart a member, the group sets any volumes with data on the member offline (for 15-20 seconds), and fails over to the secondary controller. The group sets the volumes online when the restart completes. Restarting a member has no effect on member, volume, or group configuration information or volume data stored on the member. So as long as you have the iSCSI disk timeout setup properly the host will ride out the restart.

    -Joe

    Social Media and Community Professional
    #IWork4Dell
    Get Support on Twitter - @dellcarespro

    Follow me on Twitter: @joesatdell 

  • We had an issue with one of our controllers dying and  I managed to simulate a failover by restarting as above and it works perfectly , but i have a question , how come when i go on the networks tab , i cant see the management and ISCSI IP addresses like the other 2 arrays that we have 

  • Hello, 

     Usually this points to an issue getting data from the array to the GUI.   As a check you can go to the CLI and see if that's the case.  

    GroupName>member select MEMBERNAME show eths   

     That will show you how the network ports are configured. 

    If the output doesn't show there or is very sluggish I would suggest opening a support case.  I have seen this when an array is extremely busy. 

     I would also consider upgrading the firmware. v7.1.5 was released in April 2015.  There have been many improvements since then. 

     Regards, 

    Don 

     Rather than add to an old thread, it's better to start a new one.  Especially when your issue isn't the same as the thread.  

    Social Media and Community Professional
    #IWork4Dell
    Get Support on Twitter - @dellcarespro